Does the MPG computer error change as the miles pile up?

[Iowa4Runner]

I'm working on a correction method for the MPG computer and I'd like to find out from the '03 and '04 folks if the error remains constant as miles accumulate. Has anyone with 25,000+ miles been keeping track of their mileage and how it differs from the computer? How about driving habits, does towing or off-roading seem to change the error?

 

[mason4runner]

I've only got 6k on mine, but the computer mpg is consistently 12% above the actual. It remains at 12% towing or not, hard driving or easy. The range is also not acurate. It seems to base it off previous driving, After I tow, the range is at around 200 & slowly builds back up until I tow again.
If you come up with a fix, you'll be loved!!

 

[Vortec4800]

I've only got 6k on mine, but the computer mpg is consistently 12% above the actual. It remains at 12% towing or not, hard driving or easy. The range is also not acurate. It seems to base it off previous driving, After I tow, the range is at around 200 & slowly builds back up until I tow again.

Both the MPG and the range are based off average numbers since the last time you reset the computer. If you tow, that brings down the average - and it takes a lot of normal driving to bring it back up. After you're done towing (or right after getting fuel for best results) reset the computer with the reset button and you will get a more accurate number.

This doesn't fix the computer error, but it does keep things more consistant.

 

[rando]

After each fill up, I write down my actual mileage, trip-odometer mileage, and all of the trip computer values. I then reset the computer and trip-odometer. I've only gone 4k miles in mine and most of that is city driving. What I don't have is a characterization of the type of driving I do between fill ups. That data would be only be marginally helpful in my opinion. If you want, I can compile and post what I have.

Can you tell us more about what kind of 'method' you mean? Do you plan to build some sort of pulse divider/multiplier for the speed sensor signal that feeds the trip computer? I've been planning to hook this signal to my carpc as well as whatever is in the car with respect to measuring fuel consumption. My goal was it to build a better trip computer or at least provide some more interesting statistics.

For instance, I want the computer to keep track of my mileage and maintenance history intervals and remind me when it's time to do a factory scheduled service. I'd like it to keep track of the remaining fuel and integrate with the Navigation system to alert me when it's time to refuel and automatically plot a course to the nearest Chevron station. It could warn me when begin to approach the point of no return (i.e. will run out of gas before I can reach another gas station). There are several other possibilities here.

...oops I didn't mean to hijack your thread.

 

[Good Times]

I"m at around 40k and the computer is roughly 10-11% off in general.

 

[tkr128]

I have a little over 20k miles on mine and so far the error seems to be consistent. I've been driving more in the city lately and while the mileage goes down, the difference between actual and "computed" values seems to be the same, about 3. I don't write down the number from the computer, so I don't know how useful I can be. I just look at the reading once before I reset it and compare it to my calculated values later. Also, as the MPG gets lower on the computer, it gets closer to the actual value and as it goes higher, the more error there is. I've never towed or been off-road so I can't tell you if that affects it. I also reset everything each time I fill-up.

 

[HuntRunner]

Has anyone ever asked and/or complained to Toyota about this? What was their response? Obviously there was not a solution........at that time anyway.

 

[DaveS]

You would think that since the calculation is simply based on miles travelled and fuel consumed that a simple change to the algorithm inside the computer would fix this - ie an ECU flash (with all the fun that entails of course...)

 

[mswscco]

Ha, my wife (girlfriend/fiance at that time) used to roll her eyes when I'd get out the book to record gas and mileage at fill-ups and wouldn't believe I wasn't the only person in the world to do it.

She drives and gases up the 4runner so I don't know what our calculated mileage is. Troubling thing is that the display has been at 16.? for the past few months. V6, very little 4WD during that time. I haven't heard of any V6's getting that low...

 

[4RunnerBubba]

It is possible that the computer doesn't take into consideration when you start your engine or stand in idle. These would consume gas but register no miles.

Might explain why the experience mention above about the only time its consistent is if 4-r is run consistently without any restarts and no stops.

 

[slowginjoe]

anybody use the maintenance reminder feature of the navigation unit? i haven't taken the time to mess with it yet. there's so many reminders on there (air filter, oil, trans.etc...). since the mpg computer seems to be off, i was just wondering if the maintenance calculations were accurate. i sure hope so since it's purely based on mileage instead of mileage and gas consumed..

 

[Iowa4Runner]

Can you tell us more about what kind of 'method' you mean? Do you plan to build some sort of pulse divider/multiplier for the speed sensor signal that feeds the trip computer?

Close... Instead of changing the sensed speed, my idea involves tricking the trip computer into thinking the engine is using more fuel. The trip computer monitors the #1 fuel injector open collector output from the ECM and uses the pulse length to determine fuel consumption. See pages 160, 1730, and 1731 in section 5 of the '04 service manual. My plan is to use a PIC12F683 microcontroller to alter the pulse width signal on it's way to the trip computer. There are analog circuits that might work, but the one I designed didn't perform very well when I simulated it in Pspice.

I took a test drive the other night with my oscilloscope riding shotgun so I could measure the min and max fuel injector pulse length. It was quite a challenge, but I was able to drive and glance at the scope well enough to see that the approximate range is 1.5 to 25 milliseconds. If I actually get brave enough to try this, I'll start by simply adding 10-12% to the pulse length and see how that does for a few tanks. I don't know how well pulse length and fuel flow correspond and I don't know how "smart" Toyota's fuel consumption algorithm is, so there's no telling how well my idea will work.

Issues of concern- 1) It may not consistently correct the error. 2) I really don't want to butcher my wiring harness. 3) If the ECM to fuel injector wire gets shorted to ground, it could be the end of my 4Runner. Worse case I figure I'll end up with a melted catalytic converter, a broken #1 connecting rod, and a huge fire. I don't know how likely any of that is, but just the thought makes me nervous.

I'll let you know if I get anywhere with this... :nerd:

 

[khaug]

I've carefully calculated my fuel mileage (including determining and correcting for odometer error - 2.5% on my vehicle), and I've concluded that the mpg error is neither linear nor consistent. I've long since given up using the average mpg figure for anything other than the most general indication of how much fuel I'm burning.

If you're even fussier than I and as capable with electronics as you seem to be, I think you'd do better to build something from scratch! To answer your original question, we now have 35K + miles on our '03 and I don't think the error has changed perceptibly.

Good luck.

 

[Only4Gold]

25K in 11 months...and counting

I agreee with the other posters...my computer MPG has remained constant over the 25,000 great miles I've logged so far, about 10% off reality

 

[CMiYC]

I took a test drive the other night with my oscilloscope riding shotgun so I could measure the min and max fuel injector pulse length.

Maybe I'm missing something there but if you're just measuring the injector's pulse width, why do you have to drive anywhere?

Why not just correlate the RPM to what the pulse length is? You can do that sitting in your driveway.

 

[rando]

Close... Instead of changing the sensed speed, my idea involves tricking the trip computer into thinking the engine is using more fuel.

Yah, obviously I knew there was this other half of the equation. I didn't bring it up because I had yet to come across how it was done in the manuals. Thanks for the reference.

I took a test drive the other night with my oscilloscope riding shotgun so I could measure the min and max fuel injector pulse length.

OMG LMAO! I thought for sure I was the only one ever to attempt such lunacy. My g/f was laughing at me just a few weeks ago when I made her drive while I monitored the speed sensor pulses.

Issues of concern- 1) It may not consistently correct the error.

Not really a concern. Worst case you are closer to the actual mpg than what you have now.

2) I really don't want to butcher my wiring harness.

You can add and remove wires/pins from the connectors. DONT CUT WIRES....NOOOOO.

3) If the ECM to fuel injector wire gets shorted to ground, it could be the end of my 4Runner. Worse case I figure I'll end up with a melted catalytic converter, a broken #1 connecting rod, and a huge fire. I don't know how likely any of that is, but just the thought makes me nervous.

I know the feeling. I'm pretty sure dicking up the speed sensor is certain to cause all kinds of chaos. Frying any of the major ECUs is a bad bad thing too. If you do your simulations and then build a proto board for final testing, you should be able to do the install safely without grounding anything out. Just make sure you build in protection circuits in case something fries so there is a path for the current to go other than through your ECU(s). Obviously your point about leaving the injector on continuously is really really bad.

Let me know if you need any help.

 

[rando]

See pages 160, 1730, and 1731 in section 5 of the '04 service manual.

One other word of caution. Don't trust the manuals when it comes to signal levels and such. For instance, there are several places in there where it defines the speed sensor signal as a 5v pk-pk square wave. It's actually at alternator/battery level (i.e. 13.8V pk-pk).

 

[alpha]

One other word of caution. Don't trust the manuals when it comes to signal levels and such. For instance, there are several places in there where it defines the speed sensor signal as a 5v pk-pk square wave. It's actually at alternator/battery level (i.e. 13.8V pk-pk).

wow, that's quite a difference. i'm surprised to see errors in the manuals. i suppose it's not that surprising, but still..

 

[rando]

Yep ... then again, I might be wrong. Dont' trust me either. Check everything yourself. Toyota won't be liable for any mistakes in there manual. I'm not liable for any mistakes I make either.

Another interesting tidbit...has anyone found any information at all in the service manuals about the trip computer? As far as I can tell, it was completely omitted from the wiring manul. Sure it mentions the A/C controls and the security indcator light but never the trip computer. That's why I never discovered the signal tapping the #1 injector. It only shows it going to the combination meter but I assume it's also on the D3 plug behind the trip computer, right?

 

[Iowa4Runner]

Maybe I'm missing something there but if you're just measuring the injector's pulse width, why do you have to drive anywhere?

Why not just correlate the RPM to what the pulse length is? You can do that sitting in your driveway.

I needed to measure the length of time the injector is open (duty cycle), not the frequency of injections. I had to drive it so I could monitor injector duration during the min and max fuel flow situations (coasting down a hill and full throttle acceleration).